Skip to content

fix(proofs): take the template's check-proofs.sh - #25

Merged
hyperpolymath merged 1 commit into
mainfrom
fix/proof-gate-sync
Nov 15, 2026
Merged

hyperpolymath merged 1 commit into
mainfrom
fix/proof-gate-sync

Conversation

@hyperpolymath

Copy link
Copy Markdown
Owner

estate-rules also fails on atoma main at check_proofs_test (9 of 14 controls), because atoma's scripts/check-proofs.sh is an older version than the template's. This takes the template's version.

What the template's version adds (holes first)

lean sorry handling; coqc exiting 0 on Admitted (caught via scripts/scan-dangerous.sh, already identical here); idris2 unsolved holes; and the PROOF-POLICY §2.1 Agda invocation (--cubical-compatible).

Checked, not assumed: atoma's extra scan roots

Atoma's copy hard-coded two extra roots (verification/atoma, examples/atoma). The template scans the proof dir plus every root the MANIFESTs name. Atoma's agda and idris2 MANIFESTs already name verification/atoma/agda and examples/atoma/frame-v1/abi, and all 4 .agda and 3 .idr files under those trees sit under the named roots, so both scripts cover the same files today. Trade-off: a future proof file added elsewhere under verification/atoma or examples/atoma is no longer flagged as unlisted unless a manifest names its root.

Verified locally

check_proofs_test rc 0 (was failing); the real gate still fails, not skips, with no toolchain (rc 1). With #24 also applied, 57 of the 58 plain estate-rules steps pass locally; the other, chain_mint_test, needs uuidgen (absent in my sandbox) and passes in CI on apodeixis (same step, estate-rules green on its main).

NOT verified (please read)

Agda, Idris2, Coq and Lean are not installed in my sandbox. Whether atoma's real proofs pass the stricter gate and the added --cubical-compatible flag is shown only by CI. If they do not, that is a finding about the proofs to fix, not a reason to loosen the gate.

Decision I made for you

I asked this as a selector; it was interrupted, so I took the option I had marked recommended. Close this PR to keep atoma on its older, weaker gate. This PR alone does not clear estate-rules: it needs #24 as well (the PROOF-STATUS shape).


Generated by Claude Code

check_proofs_test (estate-rules) fails on atoma main (9 of 14 controls) because
atoma's scripts/check-proofs.sh is an older version than the template's. The
template's closes gate holes: lean 'sorry', coqc exiting 0 on Admitted (via
scripts/scan-dangerous.sh, already identical here), idris2 unsolved holes, and
the PROOF-POLICY section 2.1 Agda invocation (--cubical-compatible).

Atoma's copy also hard-coded two extra scan roots (verification/atoma,
examples/atoma). Checked, not assumed: atoma's agda and idris2 MANIFESTs already
name verification/atoma/agda and examples/atoma/frame-v1/abi, and all 4 .agda
and 3 .idr files under those trees sit under the named roots, so the template's
manifest-named-roots scan covers the same files today.

Verified locally: check_proofs_test rc 0 (was failing); the gate still fails,
not skips, with no toolchain (rc 1); bash -n ok.
NOT verified: Agda, Idris2, Coq and Lean are not installed here, so whether
atoma's real proofs pass the stricter gate and the added --cubical-compatible
flag is only shown by CI. If they do not, that is a finding to fix in the
proofs, not a reason to loosen the gate.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0136P7Zk1VxiWdKq68AQx2Ev
@github-actions

github-actions Bot commented Nov 15, 2026 •

Copy link
Copy Markdown

Hypatia grid green

🟢 clean · commit 3f0e110b1092 · run 43710677988 · scope: this change (1 file(s))

Advisory: 11 row(s), worst 🟠 indeterminate (0 critical, 0 high in grid scope) — reported, never grading the change.

Unique graded findings across gating and advisory rows: 1 (1 reporter observations; 1 without sufficient identity to corroborate). Row counts remain per reporter; unique totals do not change gating.
Raw Hypatia alert-derived observations: 36 (code-scanning, Dependabot and secret-scanning; some may also be in SARIF; not an extra total).

Reporter Gate Status Critical High Medium Minor Since previous revision
Hypatia scan gating 🟢 clean 0 0 0 0 — same commit (no revision comparison)
panic-attack assail advisory 🟡 medium 0 0 1 0 — same commit (no revision comparison)
pons asinorum advisory 🟢 clean 0 0 0 0 — same commit (no revision comparison)
echidnabot advisory ⚪ not applicable 0 0 0 0 — same commit (no revision comparison)
rhodibot (RSR compliance) advisory 🟠 indeterminate 0 0 0 0 — same commit (no revision comparison)
panicbot advisory ⚪ not applicable 0 0 0 0 — same commit (no revision comparison)
sustainabot advisory ⚪ not applicable 0 0 0 0 — same commit (no revision comparison)
glambot (presentation quality) advisory 🟠 indeterminate 0 0 0 0 — same commit (no revision comparison)
seambot advisory ⚪ not applicable 0 0 0 0 — same commit (no revision comparison)
finishingbot (release readiness) advisory 🟠 indeterminate 0 0 0 0 — same commit (no revision comparison)
cipherbot advisory ⚪ not applicable 0 0 0 0 — same commit (no revision comparison)
accessibilitybot advisory ⚪ not applicable 0 0 0 0 — same commit (no revision comparison)
Hypatia scan: 4 finding(s)
  • 4 further finding(s) in files this change does not touch: reported, not graded (0 critical, 3 high).
panic-attack assail: 20 finding(s)
  • medium CommandInjection — scripts/check-proofs.sh — 33 potentially unquoted variable expansions in scripts/check-proofs.sh
  • 19 further finding(s) in files this change does not touch: reported, not graded (0 critical, 2 high).
pons asinorum: 9 finding(s)
  • 9 further finding(s) in files this change does not touch: reported, not graded (0 critical, 0 high).
echidnabot: not_applicable — template default: enrol once the repository carries proofs and an ECHIDNA engine is configured
  • ⚪ not applicable: template default: enrol once the repository carries proofs and an ECHIDNA engine is configured
rhodibot (RSR compliance): not_run — enrolled, but no report-back was received
  • 🟠 indeterminate: enrolled, but no report-back was received
panicbot: not_applicable — panic-attack is already graded by the verdict grid's own panic-attack row, from static-analysis-gate.yml evidence; enrolling panicbot would report one scanner twice and give two rows that can disagree
  • ⚪ not applicable: panic-attack is already graded by the verdict grid's own panic-attack row, from static-analysis-gate.yml evidence; enrolling panicbot would report one scanner twice and give two rows that can disagree
sustainabot: not_applicable — reserved fleet slot; no adapter exists (gitbot-fleet bots/sustainabot/README.adoc)
  • ⚪ not applicable: reserved fleet slot; no adapter exists (gitbot-fleet bots/sustainabot/README.adoc)
glambot (presentation quality): not_run — enrolled, but no report-back was received
  • 🟠 indeterminate: enrolled, but no report-back was received
seambot: not_applicable — template default: enrol once a seam register exists at spec/seams/seam-register.json
  • ⚪ not applicable: template default: enrol once a seam register exists at spec/seams/seam-register.json
finishingbot (release readiness): not_run — enrolled, but no report-back was received
  • 🟠 indeterminate: enrolled, but no report-back was received
cipherbot: not_applicable — template default: enrol once the repository implements cryptography
  • ⚪ not applicable: template default: enrol once the repository implements cryptography
accessibilitybot: not_applicable — template default: enrol once the repository serves HTML
  • ⚪ not applicable: template default: enrol once the repository serves HTML

Scope: static findings reported by each row's tool at the named commit. The table grades only findings in files this change touches, so the colour answers "does this change introduce anything?", not "what does the whole repository still carry?". Standing findings are counted in each accordion and in report.json; the baseline filter remains the place where acknowledged debt is recorded. The lane state is the worst GATING row; advisory rows are listed and counted but never set it. Purple, red, yellow and blue grade the worst GRADED finding present; orange is evidence missing, malformed or a scanner that did not run; grey is an explicit scope exclusion with its reason. No row is a security certification. report.json, badge.json and the raw findings are in the verdict-grid artifact on the linked run.

@sonarqubecloud

Copy link
Copy Markdown

Repository owner deleted a comment from coderabbitai Bot Nov 15, 2026
@hyperpolymath
hyperpolymath marked this pull request as ready for review November 15, 2026 11:49
@hyperpolymath
hyperpolymath enabled auto-merge (squash) November 15, 2026 11:49
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Nov 15, 2026 •

Copy link
Copy Markdown
Codex Review Summary: ✅ Completed

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-11-15T11:52:24.321848Z 3f0e110 Draft marked ready
🔒 Security Review ✅ Completed 2026-11-15T11:55:01.232836Z 3f0e110 Draft marked ready
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

Repository owner deleted a comment from coderabbitai Bot Nov 15, 2026
@hyperpolymath
hyperpolymath merged commit bd75d44 into main Nov 15, 2026
49 of 50 checks passed
@hyperpolymath
hyperpolymath deleted the fix/proof-gate-sync branch November 15, 2026 11:52

Copy link
Copy Markdown
Owner Author

This branch now also contains the PROOF-STATUS shape fix from #24 (merged in as 14c1056, signed). Reason: each PR alone leaves estate-rules red. #24 fixes the claims reader and the vocabulary test but then stops at the proof-gate self-test; this PR fixes the proof gate but sits behind the claims reader. So #25 on its own could never go green, and auto-merge here would wait forever. With both, 57 of 58 estate-rules steps pass locally (the 58th, chain_mint_test, needs uuidgen, which my sandbox lacks; it passes in CI on apodeixis).

Still not verified: whether atoma's real proofs pass the stricter check-proofs.sh (Agda/Idris2/Coq/Lean are not installed in my sandbox). This PR's CI is now the first real test. If it fails there, that is a finding about the proofs, not a reason to loosen the gate.

If #24 merges first, this PR just shrinks to the check-proofs.sh change; nothing conflicts.


Generated by Claude Code

Copy link
Copy Markdown
Owner Author

Correction to my previous comment: this PR was already merged (bd75d44, the 3f0e110 version, via auto-merge) before I pushed the extra commit 14c1056 to its branch. That commit never reached main and is not needed: it only duplicated #24's change. Please ignore it, and ignore my claim that auto-merge "would wait forever" — it did not, because estate-rules is evidently not a required check.

What this means now: main has the template's check-proofs.sh (from this PR) but still lacks the PROOF-STATUS shape fix, so estate-rules on main stays red at the claims reader until #24 merges. #24 is rebased on the new main and its checks are running. I should have re-checked this PR's state before pushing.


Generated by Claude Code

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants